Re: [patch] tux3-2.6.0-D0

Ingo Molnar <[email protected]>
Newsgroups gmane.network.tux
Message-ID <[email protected]>
ah - do you have CONFIG_PREEMPT on? If yes then please disable it.

	Ingo

* joe <[email protected]> wrote:

> Ingo Molnar wrote:
> 
> >* joe <[email protected]> wrote:
> >
> > 
> >
> >>>	redhat.com/~mingo/TUX-patches/tux3-2.6.0-D0
> >>>
> >>>this should fix the compilation problems.
> >>>
> >>>
> >>>     
> >>>
> >>hmm, it builds and runs, but seems to be considerably slower than apache 
> >>:(
> >>   
> >>
> >
> >via what utility did you test them, and over what kind of network?
> >
> > 
> >
> I tested via apachebench, both from localhost and from a 100 MB lan 
> connection -
> 
> I'm getting about 500 connections/sec with apache 2, and 50 
> connections/sec with tux3
> 
> The tux that comes with redhat 9 will do 15000-20000 connections/sec on 
> the same server, so I know something is obviously not right with this 
> whole setup. The server seems quite snappy otherwise, subjectively speaking.
> 
> Linux jyro.mirai.cx 2.6.1-rc1-tux #1 Thu Jan 1 12:08:40 PST 2004 i686 
> i686 i386 GNU/Linux
> 
> Hmm, I just noticed a bunch of these in the syslog - perhaps a useful clue?
> 
> bad: scheduling while atomic!
> Call Trace:
> [<c0120190>] schedule+0x8e0/0x8f0
> [<e0ec4381>] process_requests+0x241/0x2e0 [tux]
> [<e0ed2b76>] event_loop+0x106/0x1c0 [tux]
> [<e0ed4c02>] __sys_tux+0x262/0x13a0 [tux]
> [<c011e868>] kernel_map_pages+0x28/0x70
> [<c0195b00>] dput+0x30/0x720
> [<c0189366>] path_release+0x16/0x40
> [<c0172a18>] sys_chdir+0x58/0x90
> [<c034a81d>] sys_tux+0x3bd/0x3d0
> [<c010a3eb>] syscall_call+0x7/0xb
> 
> bad: scheduling while atomic!
> Call Trace:
> [<c0120190>] schedule+0x8e0/0x8f0
> [<e0ec42a4>] process_requests+0x164/0x2e0 [tux]
> [<e0ed2b76>] event_loop+0x106/0x1c0 [tux]
> [<e0ed4c02>] __sys_tux+0x262/0x13a0 [tux]
> [<c011e868>] kernel_map_pages+0x28/0x70
> [<c0195b00>] dput+0x30/0x720
> [<c0189366>] path_release+0x16/0x40
> [<c0172a18>] sys_chdir+0x58/0x90
> [<c034a81d>] sys_tux+0x3bd/0x3d0
> [<c010a3eb>] syscall_call+0x7/0xb
> 
> bad: scheduling while atomic!
> Call Trace:
> [<c0120190>] schedule+0x8e0/0x8f0
> [<e0ec42cc>] process_requests+0x18c/0x2e0 [tux]
> [<e0ed2b76>] event_loop+0x106/0x1c0 [tux]
> [<e0ed4c02>] __sys_tux+0x262/0x13a0 [tux]
> [<c011e868>] kernel_map_pages+0x28/0x70
> [<c0195b00>] dput+0x30/0x720
> [<c0189366>] path_release+0x16/0x40
> [<c0172a18>] sys_chdir+0x58/0x90
> [<c034a81d>] sys_tux+0x3bd/0x3d0
> [<c010a3eb>] syscall_call+0x7/0xb
> 
> 
> Joe
>
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.